3 ms·

This is an excellent post about the evolution and the milestones that lead to the current async Rust approach. I thought the technical arguments were clearly ar

This is an excellent post about the evolution and the milestones that lead to the current async Rust approach. I thought the technical arguments were clearly articulated, dispassionately responding to the recent critical viewpoints regarding async Rust. I hope that people, especially the ones that have voiced these criticisms, take the time to read and digest this article. It may not change their mind, but perhaps it will help them understand the current situation better. Unfortunately, there are already some comments that prove my optimism misguided elsewhere in this thread.

soerxpso 3y ago
Although the articles arguments show that the decisions made were understandable and perhaps the "right decisions" for the language, there's a common theme that backwards compatibility was a motivator. I'm not saying they should have broken backwards compatibility, but "This is actually the best way to do it," and, "This is the best way that we could do it given our tough situation," are two very different points. I don't really care about the latter when I'm evaluating whether I actually like the way it's done.
- withoutboats3 3y ago
  The only thing constrained by backward compatibility was using Pin instead of Move. Async/await itself is the natural progression for Rust for the reasons I outlined at length regarding how Rust represents zero-cost coroutines.